how to make a rolling mean in pandas but only for items that have the same id/value?

Question

columns: datetime | clientid | amounts | *new_column_to_be_implemented* (rolling mean of values before but only for values that are the same in clientid)

<